MySQL Forums
Forum List  »  NDB clusters

Re: ndbd's stuck in phase 2
Posted by: Mikael Ronström
Date: June 21, 2016 04:35PM

You have specified in your config that you want to have
100000 FragmentLogfiles. The default is that a REDO log
file is 64 MByte and there are by default 4 log parts.
So this means that every log file will 256MByte and you
have specified that you want 100000 of those. This means
that you will have 25.6 TByte of REDO log.

So stop the start, remove the ndb_2_fs folders and
decrease FragmentLogfiles to a more reasonable value.

The reason the restart is stuck is simply that we always
initialise the REDO log as part of an initial start to
ensure that we have allocated the disk space before we
start (otherwise we might run out of REDO log instead
while executing).

Writing 25.6 TByte takes some time and I presume you
don't have sufficient disk space for it either.

Rgrds Mikael ronstrom

Options: ReplyQuote

Written By
June 20, 2016 02:37PM
June 21, 2016 01:53AM
June 21, 2016 04:04AM
Re: ndbd's stuck in phase 2
June 21, 2016 04:35PM

Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.